Add Guide to Pharmacology

[jsstevenson]

eg https://bioregistry.io/registry/iuphar.ligand, https://bioregistry.io/registry/iuphar.receptor

I used iuphar as a catch-all bioregistry ID -- since it's causing one of the tests to fail, should I use one of the subdomains instead (we only need iuphar.ligand for our in-house projects) or just leave it out?

[cthoyt]

Bringing up the difference between Guide to Pharmacology, the database, vs. the ligand and receptor vocabularies, is a really interesting thought.

TL;DR - no bioregistry ID needed here.

I don't have a good resource to point to at the moment (in fact, I'm writing up a glossary to go with the Bioregistry and Bioversions right now) but my current thinking is that some databases/resources correspond 1-to-1 with a vocabulary, like many ontologies, but some database house multiple nomenclatures like Guide to Pharmacology. Some that have multiple nomenclatures use a single unified endpoint, like KEGG or ChEMBL, for resolving all of its many entity types, so it gets a parent prefix and then we curate a "part of" relationship between them like chembl.compound part_of chembl. Some databases have multiple nomenclatures, like Guide to Pharmacology, but not a unified endpoint, so there isn't really a good reason to have a parent prefix. Interestingly, some databases with multiple nomenclatures have one obvious one then one or more secondary nomenclatures, like HGNC, which also has a gene group nomenclature.

More discussion on this is also happening at biopragmatics/bioregistry#133